Additional PPA Agreements Reached

CAISO Net Negative Emissions signal moments when greenhouse gas intensity of serving ISO demand drops below zero, driven by high renewable generation, low load, strong solar exports, and imports accounting in the California grid.

Key Points

Moments when CAISO's CO2 to serve demand is below zero, driven by renewables, exports, and import accounting.

✅ Calculated using imports and exports to serve ISO demand

✅ Occur during high solar output, low weekend load

✅ Coincide with curtailment and record renewable penetration

We’re a long way from the land of milk and honey, but on Easter Sunday – for about an hour – we got a taste.

On Sunday, at 1:55 PM Pacific Time the California Independent Systems Operator (CAISO) reported that greenhouse gas emissions necessary to serve its demand (~80% of California’s electricity demand on an annual basis), was measured at a rate -16 metric tons of CO2 per hour. Five minutes later, the value was -2 mTCO2/h, before it crept back up to 40 mTCO2/h at 2:05 PM PST. At 2:10 PST though it fell back to -86 mTCO2/h and stayed negative until 3:05 PM PST, even as global CO2 emissions flatlined in 2019 according to the IEA.

This information was brought to the attention of pv magazine via tweet from eagle eye Jon Pa after CAISO’s site first noted the negative values:

The region was still generating CO2 though, as natural gas, biogas, biomass, geothermal and even coal plants were running and pumping out emissions, even as potent greenhouse gases declined in the US under control efforts. CAISO’s Greenhouse Gas Emission Tracking Methodology, December 28, 2016 (pdf) notes the below calculations to create the value what it terms, “Total GHG emissions to serve ISO demand”:

Of importance to note is that to get to the net negative value, CAISO considered all electricity imports and exports, a reminder that climate policy shapes grid operations across North America. And as can be noted in the image below the CO2 intensity of imports during the day rapidly declined as the sun came up, first going negative around 9:05 AM PST, and mostly staying so until just before 6 PM PST.

During this same weekend, other records were noted (reiterating that we’re in record setting season and as the state pursues its 100% carbon-free mandate now in law) such as a new electricity export record of greater than 2 GW and total renewable electricity as part of total demand at greater than 70%.

At the peak negative moment of 2:15 PM PST, -112 mTCO2/h seen below, the total amount of clean instantaneous generation being used in the power grid region was 17 GW, a far cry from heat-driven reliability strains like rolling blackout warnings that arise during extreme demand, with renewables giving 76% of the total, hydro 14%, nuclear 13% and imports of -12% countering the CO2 coming from just over 1.4 GW of gas generation.

Also of importance are a few layers of nuance in the electricity demand charts. First off we’re in the shoulder seasons  of California – nice cool weather before the warmth of summer drives air conditioning demand. Additional the weekend electricity demand is always lower, as well, Easter Sunday might have had an affect, whereas in colder regions Calgary’s electricity use can soar during frigid snaps.

Lastly to note was the amount of electricity from solar and wind generation being curtailed. And while the Sunday numbers weren’t available yet, the below image noted Saturday with 10 GWh in total being curtailed (pdf) – peaking at over 3.2 GW of instantaneous mostly solar power even as solar is now the cheapest electricity according to the IEA, in the hours of 2 and 3 PM PST. On an annualized basis, less than 2% of total potential solar electricity was curtailed in 2018.

Related News

• Electricity Forum News

• Climate Change News

U.S. power grid cyberattacks expose critical infrastructure to Russian hackers, DHS warns, targeting SCADA, smart grid sensors, and utilities; NERC CIP defenses, microgrids, and resilience planning aim to mitigate outages and supply chain disruptions.

Key Points

U.S. power grid cyberattacks target utility control systems, risking outages, disruption, requiring stronger defenses.

✅ Russian access to utilities and SCADA raises outage risk

✅ NERC CIP, DHS, and utilities expand cyber defenses

✅ Microgrids and renewables enhance resilience, islanding capability

The U.S. Department of Homeland Security has revealed that Russian government hackers accessed control rooms at hundreds of U.S. electrical utility companies, gaining far more access to the operations of many more companies than previously disclosed by federal officials.

Securing the electrical grid, upon which is built almost the entirety of modern society, is a monumental challenge. Several experts have explained aspects of the task, potential solutions and the risks of failure for The Conversation:

1. What’s at stake?

The scale of disruption would depend, in part, on how much damage the attackers wanted to do. But a major cyberattack on the electricity grid could send surges through the grid, much as solar storms have done.

Those events, explains Rochester Institute of Technology space weather scholar Roger Dube, cause power surges, damaging transmission equipment. One solar storm in March 1989, he writes, left “6 million people without power for nine hours … [and] destroyed a large transformer at a New Jersey nuclear plant. Even though a spare transformer was nearby, it still took six months to remove and replace the melted unit.”

More serious attacks, like larger solar storms, could knock out manufacturing plants that build replacement electrical equipment, gas pumps to fuel trucks to deliver the material and even “the machinery that extracts oil from the ground and refines it into usable fuel. … Even systems that seem non-technological, like public water supplies, would shut down: Their pumps and purification systems need electricity.”

In the most severe cases, with fuel-starved transportation stalled and other basic infrastructure not working, “[p]eople in developed countries would find themselves with no running water, no sewage systems, no refrigerated food, and no way to get any food or other necessities transported from far away. People in places with more basic economies would also be without needed supplies from afar.”

2. It wouldn’t be the first time

Russia has penetrated other countries’ electricity grids in the past, and used its access to do real damage. In the middle of winter 2015, for instance, a Russian cyberattack shut off the power to Ukraine’s capital in the middle of winter 2015.

Power grid scholar Michael McElfresh at Santa Clara University discusses what happened to cause hundreds of thousands of Ukrainians to lose power for several hours, and notes that U.S. utilities use software similar to their Ukrainian counterparts – and therefore share the same vulnerabilities.

3. Security work is ongoing

These threats aren’t new, write grid security experts Manimaran Govindarasu from Iowa State and Adam Hahn from Washington State University. There are a lot of people planning defenses, including the U.S. government, as substation attacks are growing across the country. And the “North American Electric Reliability Corporation, which oversees the grid in the U.S. and Canada, has rules … for how electric companies must protect the power grid both physically and electronically.” The group holds training exercises in which utility companies practice responding to attacks.

4. There are more vulnerabilities now

Grid researcher McElfresh also explains that the grid is increasingly complex, with with thousands of companies responsible for different aspects of generating, transmission, and delivery to customers. In addition, new technologies have led companies to incorporate more sensors and other “smart grid” technologies. He describes how that, as a recent power grid report card underscores, “has created many more access points for penetrating into the grid computer systems.”

5. It’s time to ramp up efforts

The depth of access and potential control over electrical systems means there has never been a better time than right now to step up grid security amid a renewed focus on protecting the grid among policymakers and utilities, writes public-utility researcher Theodore Kury at the University of Florida. He notes that many of those efforts may also help protect the grid from storm damage and other disasters.

6. A possible solution could be smaller grids

One protective effort was identified by electrical engineer Joshua Pearce at Michigan Technological University, who has studied ways to protect electricity supplies to U.S. military bases both within the country and abroad. He found that the Pentagon has already begun testing systems, as the military ramps up preparation for major grid hacks, that combine solar-panel arrays with large-capacity batteries. “The equipment is connected together – and to buildings it serves – in what is called a ‘microgrid,’ which is normally connected to the regular commercial power grid but can be disconnected and become self-sustaining when disaster strikes.”

He found that microgrid systems could make military bases more resilient in the face of cyberattacks, criminals or terrorists and natural disasters – and even help the military “generate all of its electricity from distributed renewable sources by 2025 … which would provide energy reliability and decrease costs, [and] largely eliminate a major group of very real threats to national security.”

Related News

• Electricity Forum News

• Grid Technologies News

US Electricity Demand Stagnation reflects decoupling from GDP as TVA's IRP revises outlook, with energy efficiency, distributed generation, renewables, and cheap natural gas undercutting coal, reshaping utility business models and accelerating grid modernization.

Key Points

US electricity demand stagnation is flat load growth driven by efficiency, DG, and decoupling from GDP.

✅ Flat sales pressure IOU profits and legacy baseload investments.

✅ Efficiency and rooftop solar reduce load growth and capacity needs.

✅ Utilities must pivot to services, DER orchestration, and grid software.

The US electricity sector is in a period of unprecedented change and turmoil, with emerging utility trends reshaping strategies across the industry today. Renewable energy prices are falling like crazy. Natural gas production continues its extraordinary surge. Coal, the golden child of the current administration, is headed down the tubes.

In all that bedlam, it’s easy to lose sight of an equally important (if less sexy) trend: Demand for electricity is stagnant.

Thanks to a combination of greater energy efficiency, outsourcing of heavy industry, and customers generating their own power on site, demand for utility power has been flat for 10 years, with COVID-19 electricity demand underscoring recent variability and long-run stagnation, and most forecasts expect it to stay that way. The die was cast around 1998, when GDP growth and electricity demand growth became “decoupled”:

This historic shift has wreaked havoc in the utility industry in ways large and small, visible and obscure. Some of that havoc is high-profile and headline-making, as in the recent requests from utilities (and attempts by the Trump administration) to bail out large coal and nuclear plants amid coal and nuclear industry disruptions affecting power markets and reliability.

Some of it, however, is unfolding in more obscure quarters. A great example recently popped up in Tennessee, where one utility is finding its 20-year forecasts rendered archaic almost as soon as they are released.

Falling demand has TVA moving up its planning process

Every five years, the Tennessee Valley Authority (TVA) — the federally owned regional planning agency that, among other things, supplies electricity to Tennessee and parts of surrounding states — develops an Integrated Resource Plan (IRP) meant to assess what it requires to meet customer needs for the next 20 years.

The last IRP, completed in 2015, anticipated that there would be no need for major new investment in baseload (coal, nuclear, and hydro) power plants; it foresaw that energy efficiency and distributed (customer-owned) energy generation would hold down demand.

Even so, TVA underestimated. Just three years later, the Times Free Press reports, “TVA now expects to sell 13 percent less power in 2027 than it did two decades earlier — the first sustained reversal in the growth of electricity usage in the 85-year history of TVA.”

TVA will sell less electricity in 10 years than it did 10 years ago. That is bonkers.

This startling shift in prospects has prompted the company to accelerate its schedule. It will now develop its next IRP a year early, in 2019.

Think for a moment about why a big utility like TVA (serving 9 million customers in seven states, with more than $11 billion in revenue) sets out to plan 20 years ahead. It is investing in extremely large and capital-intensive infrastructure like power plants and transmission lines, which cost billions of dollars and last for decades. These are not decisions to make lightly; the utility wants to be sure that they will still be needed, and will still pay off, for many years to come.

Now think for a moment about what it means for the electricity sector to be changing so fast that TVA’s projections are out of date three years after its last IRP, so much so that it needs to plunge back into the multimillion-dollar, year-long process of developing a new plan.

TVA wanted a plan for 20 years; the plan lasted three.

The utility business model is headed for a reckoning

TVA, as a government-owned, fully regulated utility, has only the goals of “low cost, informed risk, environmental responsibility, reliability, diversity of power and flexibility to meet changing market conditions,” as its planning manager told the Times Free Press. (Yes, that’s already a lot of goals!)

But investor-owned utilities (IOUs), which administer electricity for well over half of Americans, face another imperative: to make money for investors. They can’t make money selling electricity; monopoly regulations forbid it, raising questions about utility revenue models as marginal energy costs fall. Instead, they make money by earning a rate of return on investments in electrical power plants and infrastructure.

The problem is, with demand stagnant, there’s not much need for new hardware. And a drop in investment means a drop in profit. Unable to continue the steady growth that their investors have always counted on, IOUs are treading water, watching as revenues dry up

Utilities have been frantically adjusting to this new normal. The generation utilities that sell into wholesale electricity markets (also under pressure from falling power prices; thanks to natural gas and renewables, wholesale power prices are down 70 percent from 2007) have reacted by cutting costs and merging. The regulated utilities that administer local distribution grids have responded by increasing investments in those grids, including efforts to improve electricity reliability and resilience at lower cost.

But these are temporary, limited responses, not enough to stay in business in the face of long-term decline in demand. Ultimately, deeper reforms will be necessary.

As I have explained at length, the US utility sector was built around the presumption of perpetual growth. Utilities were envisioned as entities that would build the electricity infrastructure to safely and affordably meet ever-rising demand, which was seen as a fixed, external factor, outside utility control.

But demand is no longer rising. What the US needs now are utilities that can manage and accelerate that decline in demand, increasing efficiency as they shift to cleaner generation. The new electricity paradigm is to match flexible, diverse, low-carbon supply with (increasingly controllable) demand, through sophisticated real-time sensing and software.

That’s simply a different model than current utilities are designed for. To adapt, the utility business model must change. Utilities need newly defined responsibilities and new ways to make money, through services rather than new hardware. That kind of reform will require regulators, politicians, and risky experiments. Very few states — New York, California, Massachusetts, a few others — have consciously set off down that path.

Flat or declining demand is going to force the issue

Even if natural gas and renewables weren’t roiling the sector, the end of demand growth would eventually force utility reform.

To be clear: For both economic and environmental reasons, it is good that US power demand has decoupled from GDP growth. As long as we’re getting the energy services we need, we want overall demand to decline. It saves money, reduces pollution, and avoids the need for expensive infrastructure.

But the way we’ve set up utilities, they must fight that trend. Every time they are forced to invest in energy efficiency or make some allowance for distributed generation (and they must always be forced), demand for their product declines, and with it their justification to make new investments.

Only when the utility model fundamentally changes — when utilities begin to see themselves primarily as architects and managers of high-efficiency, low-emissions, multidirectional electricity systems rather than just investors in infrastructure growth — can utilities turn in earnest to the kind planning they need to be doing.

In a climate-aligned world, utilities would view the decoupling of power demand from GDP growth as cause for celebration, a sign of success. They would throw themselves into accelerating the trend.

Instead, utilities find themselves constantly surprised, caught flat-footed again and again by a trend they desperately want to believe is temporary. Unless we can collectively reorient utilities to pursue rather than fear current trends in electricity, they are headed for a grim reckoning.

Related News

• Electricity Forum News

• Energy Policy News

EV-Driven Electricity Demand Growth will reshape utilities through electrification, EV adoption, grid modernization, and ratebasing of charging, as NREL forecasts rising terawatt-hours, CAGR increases, and demand-side flexibility to manage emissions and reliability.

Key Points

Growth in power consumption fueled by EV adoption and electrification, increasing utility sales and grid investment.

✅ NREL projects 20%-38% higher U.S. load by 2050

✅ Utilities see CAGR up to 1.6% and 80 TWh/year growth

✅ Demand-side flexibility and EV charging optimize grids

Utilities have struggled with flat demand for years, but analysis by the National Renewable Energy Laboratory predicts steady growth across the next three decades — largely driven by the adoption of electric vehicles, including models like the Tesla Model 3 that are reshaping expectations.

The study considers three scenarios, a reference case and medium- and high-adoption electrification predictions. All indicate demand growth, but in the medium and high scenarios for 2050, U.S. electricity consumption increases by 20% and 38%, respectively, compared to business as usual.

Utilities could go from stagnant demand to compound annual growth rates of 1.6%, which would amount to sustained absolute growth of 80 terawatt-hours per year.

"This unprecedented absolute growth in annual electricity consumption can significantly alter supply-side infrastructure development requirements," the report says, and could challenge state power grids in multiple regions.

NREL's Trieu Mai, principal investigator for the study, cautions that more research is needed to fully assess the drivers and impacts of electrification, "as well as the role and value of demand-side flexibility."

"Although we extensively and qualitatively discuss the potential drivers and barriers behind electric technology adoption in the report, much more work is needed to quantitatively understand these factors," Mai said in a statement.

However, utilities have largely bought into the dream.

"Electric vehicles are the biggest opportunity we see right now," Energy Impact Partners CEO Hans Kobler told Utility Dive. And the impact could go beyond just higher kilowattt-hour sales, particularly as electric truck fleets come online.

"When the transportation sector is fully electrified, it will result in around $6 trillion in investment," Kobler said. "Half of that is on the infrastructure side of the utility." And the industry can also benefit through ratebasing charging stations and managing the new demand.

One benefit that NREL's report points to is the possibility of "expanded value streams enabled by electric and/or grid-connected technologies," such as energy storage and mobile chargers that enhance flexibility.

"Many electric utilities are carefully watching the trend toward electrification, as it has the potential to increase sales and revenues that have stagnated or fallen over the past decade," the report said, highlighting potential benefits for all customers as adoption grows. "Beyond power system planning, other motivations to study electrification include its potential to impact energy security, emissions, and innovation in electrical end-use technologies and overall efficient system integration. The impacts of electrification could be far-reaching and have benefits and costs to various stakeholders."

Related News

• Electricity Forum News

• EV Infrastructure News

Maine Climate Council Act targets 80% renewable power by 2030 and 100% by 2050, slashing greenhouse gas emissions via clean electricity, grid procurement, long-term contracts, wind and hydro integration, resilience planning, and carbon sequestration.

Key Points

A Maine policy forming a Climate Council to reach 80% renewables in 2030 100% in 2050 and cut greenhouse gas emissions.

✅ 80% renewable electricity by 2030; 100% by 2050.

✅ 45% GHG cut by 2030; 80% by 2050.

✅ Utility procurement authority for clean capacity and energy.

The winds of change have shifted and are blowing Northward, as Maine’s Governor, Janet T. Mills, has put forth an act establishing a Climate Council to guide the state’s consumption to 80% renewable electricity in 2030 and 100% by 2050, echoing New York's Green New Deal ambitions underway.

The act, LR 2478 (pdf), also sets a goal of reducing greenhouse gas emissions by 45% in 2030 and 80% by 2050. The document will be submitted to the state Legislature for consideration.

The commission would have the authority to direct investor owned transmission and distribution utilities to run competitive procurement processes, and enter into long-term contracts for capacity resources, energy resources, renewable energy credit contracts, and participate in regional programs, as these all lead toward the clean electricity and emissions-reducing goals that mirror California's 100% mandate debates today.

The Climate Council would convene industry working groups, including Scientific and Technical, Transportation, Coastal and Marine, Energy, and Building & Infrastructure working groups, plus others as needed, where examples like New Zealand's electricity transition could inform discussions.

Membership within the council would include two members of the State Senate, two members of the House, a tribal representative, many department commissioners (Education, Defense, Transportation, etc.), multiple directors, business representatives, environmental non-profit members, and climate science and resilience representatives as well.

The council would update the Maine State Climate Plan every four years, and solicit input from the public and report out progress on its goals every two years, similar to planning underway in Minnesota's carbon-free plan framework. The first Climate Action Plan would be submitted to the legislature by December 1, 2020.

Specifically, the responsibilities of the Scientific and Technical Subcommittee were laid out. The group would be scheduled to meet at least every six months, beginning no later than October 1, 2019. The group would be tasked with reviewing existing scientific literature, including net-zero electricity pathways research, to use it as guidance, recognizing gaps in the state’s knowledge, and guiding outside experts to ascertain this knowledge.  The group would consider ocean acidification, and climate change effects on the state’s species; establish science-based sea-level rise projections for the state’s coastal regions by December 1, 2020; create a climate risk map for flooding and extreme weather events; and consider carbon sequestration via biomass growth.

The state’s largest power plants (above image), generate about 31% from gas, 28% from wood and 41% from hydro+wind. Already, the state has a very clean electricity profile, much like efforts to decarbonize Canada's power sector continue apace. Below, the U.S. Energy Information Administration (EIA) notes that 51% of electricity generation within the state comes from mostly wind+hydro, with a small touch from solar power. The state also gets 24% from wood and other biomass, which would lead some to argue that the state is already at 75% “renewable electricity”. The Governor’s document does reference wind power specifically as a renewable, however, no other specific electricity source. And there is much reference to forestry, agriculture, and logging – specifically noting carbon sequestration – but nothing regarding electricity.

The state’s final 25% of electricity mostly comes from natural gas, even as renewable electricity momentum builds across North America, with this author choosing to put “other” under the fossil percentage noted above.

Related News

• Electricity Forum News

• Renewable Energy News

ICT Electricity Demand is surging as data centers, 5G, IoT, and server farms expand, straining grids, boosting carbon emissions, and challenging climate targets unless efficiency, renewable energy, and smarter cooling dramatically improve.

Key Points

ICT electricity demand is power used by networks, devices, and data centers across the global communications sector.

✅ Projected to reach up to 20 percent of global electricity by 2025

✅ Driven by data centers, 5G traffic, IoT, and high-res streaming

✅ Mitigation: efficiency, renewable PPAs, advanced cooling, workload shifts

The communications industry could use 20% of all the world’s electricity by 2025, hampering attempts to meet climate change targets, even as countries like New Zealand's electrification plans seek broader decarbonization, and straining grids as demand by power-hungry server farms storing digital data from billions of smartphones, tablets and internet-connected devices grows exponentially.

The industry has long argued that it can considerably reduce carbon emissions by increasing efficiency and reducing waste, but academics are challenging industry assumptions. A new paper, due to be published by US researchers later this month, will forecast that information and communications technology could create up to 3.5% of global emissions by 2020 – surpassing aviation and shipping – and up to 14% 2040, around the same proportion as the US today.

Global computing power demand from internet-connected devices, high resolution video streaming, emails, surveillance cameras and a new generation of smart TVs is increasing 20% a year, consuming roughly 3-5% of the world’s electricity in 2015, says Swedish researcher Anders Andrae.

In an update o a 2016 peer-reviewed study, Andrae found that without dramatic increases in efficiency, the ICT industry could use 20% of all electricity and emit up to 5.5% of the world’s carbon emissions by 2025. This would be more than any country, except China, India and the USA, where China's data center electricity use is drawing scrutiny.

He expects industry power demand to increase from 200-300 terawatt hours (TWh) of electricity a year now, to 1,200 or even 3,000TWh by 2025. Data centres on their own could produce 1.9 gigatonnes (Gt) (or 3.2% of the global total) of carbon emissions, he says.

“The situation is alarming,” said Andrae, who works for the Chinese communications technology firm Huawei. “We have a tsunami of data approaching. Everything which can be is being digitalised. It is a perfect storm. 5G [the fifth generation of mobile technology] is coming, IP [internet protocol] traffic is much higher than estimated, and all cars and machines, robots and artificial intelligence are being digitalised, producing huge amounts of data which is stored in data centres.”

US researchers expect power consumption to triple in the next five years as one billion more people come online in developing countries, and the “internet of things” (IoT), driverless cars, robots, video surveillance and artificial intelligence grows exponentially in rich countries.

The industry has encouraged the idea that the digital transformation of economies and large-scale energy efficiencies will slash global emissions by 20% or more, but the scale and speed of the revolution has been a surprise.

Global internet traffic will increase nearly threefold in the next five years says the latest Cisco Visual Networking Index, a leading industry tracker of internet use.

“More than one billion new internet users are expected, growing from three billion in 2015 to 4.1bn by 2020. Over the next five years global IP networks will support up to 10bn new devices and connections, increasing from 16.3bn in 2015 to 26bn by 2020,” says Cisco.

A 2016 Berkeley laboratory report for the US government estimated the country’s data centres, which held about 350m terabytes of data in 2015, could together need over 100TWh of electricity a year by 2020. This is the equivalent of about 10 large nuclear power stations.

Data centre capacity is also rocketing in Europe, where the EU's plan to double electricity use by 2050 could compound demand, and Asia with London, Frankfurt, Paris and Amsterdam expected to add nearly 200MW of consumption in 2017, or the power equivalent of a medium size power station.

“We are seeing massive growth of data centres in all regions. Trends that started in the US are now standard in Europe. Asia is taking off massively,” says Mitual Patel, head of EMEA data centre research at global investment firm CBRE.

“The volume of data being handled by such centres is growing at unprecedented rates. They are seen as a key element in the next stage of growth for the ICT industry”, says Peter Corcoran, a researcher at the university of Ireland, Galway.

Using renewable energy sounds good but no one else benefits from what will be generated, and it skews national attempts to reduce emissions

Ireland, which with Denmark is becoming a data base for the world’s biggest tech companies, has 350MW connected to data centres but this is expected to triple to over 1,000MW, or the equivalent of a nuclear power station size plant, in the next five years.

Permission has been given for a further 550MW to be connected and 750MW more is in the pipeline, says Eirgrid, the country’s main grid operator.

“If all enquiries connect, the data centre load could account for 20% of Ireland’s peak demand,” says Eirgrid in its All-Island Generation Capacity Statement 2017-2026  report.

The data will be stored in vast new one million square feet or larger “hyper-scale” server farms, which companies are now building. The scale of these farms is huge; a single $1bn Apple data centre planned for Athenry in Co Galway, expects to eventually use 300MW of electricity, or over 8% of the national capacity and more than the daily entire usage of Dublin. It will require 144 large diesel generators as back up for when the wind does not blow.

 Facebook’s Lulea data centre in Sweden, located on the edge of the Arctic circle, uses outside air for cooling rather than air conditioning and runs on hydroelectic power generated on the nearby Lule River. Photograph: David Levene for the Guardian

Pressed by Greenpeace and other environment groups, large tech companies with a public face , including Google, Facebook, Apple, Intel and Amazon, have promised to use renewable energy to power data centres. In most cases they are buying it off grid but some are planning to build solar and wind farms close to their centres.

Greenpeace IT analyst Gary Cook says only about 20% of the electricity used in the world’s data centres is so far renewable, with 80% of the power still coming from fossil fuels.

“The good news is that some companies have certainly embraced their responsibility, and are moving quite aggressively to meet their rapid growth with renewable energy. Others are just growing aggressively,” he says.

Architect David Hughes, who has challenged Apple’s new centre in Ireland, says the government should not be taken in by the promises.

“Using renewable energy sounds good but no one else benefits from what will be generated, and it skews national attempts to reduce emissions. Data centres … have eaten into any progress we made to achieving Ireland’s 40% carbon emissions reduction target. They are just adding to demand and reducing our percentage. They are getting a free ride at the Irish citizens’ expense,” says Hughes.

Eirgrid estimates indicate that by 2025, one in every 3kWh generated in Ireland could be going to a data centre, he added. “We have sleepwalked our way into a 10% increase in electricity consumption.”

Fossil fuel plants may have to be kept open longer to power other parts of the country, and manage issues like SF6 use in electrical equipment, and the costs will fall on the consumer, he says. “We will have to upgrade our grid and build more power generation both wind and backup generation for when the wind isn’t there and this all goes onto people’s bills.”

Under a best case scenario, says Andrae, there will be massive continuous improvements of power saving, as the global energy transition gathers pace, renewable energy will become the norm and the explosive growth in demand for data will slow.

But equally, he says, demand could continue to rise dramatically if the industry keeps growing at 20% a year, driverless cars each with dozens of embedded sensors, and cypto-currencies like Bitcoin which need vast amounts of computer power become mainstream.

“There is a real risk that it all gets out of control. Policy makers need to keep a close eye on this,” says Andrae.

Related News

• Electricity Forum News

• Climate Change News